Marcus Andrey Study
Ler em português Library

Motion and Interaction · Step 1 · 8 min read

When not to animate

The most useful motion skill is knowing when to leave it out. Five questions to ask before any animation, why the third time matters more than the first, and the motion that makes some people physically unwell.

Every animation costs the person a little time and a little attention. Most of the motion skill is deciding when that cost is worth paying. An animation is a promise that something changed and it matters. Break that promise often enough, with motion that only decorates, slows down or moves at the edge of the page, and people stop trusting it. For some of them, it also makes them dizzy. So before you choose a duration or a curve, ask whether it should move at all.

What you'll be able to do after reading:

  • Run the five questions before adding any animation
  • Judge an animation on the third use, not the first
  • Recognise the motion that harms people with vestibular disorders
  • Defend "no animation" as a design decision, not a shortcut

1. Five questions before you animate

  1. Does it explain a change they need to notice or understand?If not: don't animate. A cut is fine.
  2. Will they see it many times a day?If yes: keep it short, around 100 to 200 ms, or make it a fade, or nothing.
  3. Does it make them wait before they can act?If yes: let them act during it, or cut it.
  4. Does it move at the edge of the screen while they read or work?If yes: don't move it. A fade at most, or a static change.
  5. Does it still work with reduced motion on?If not: design that version, a fade or an instant change that says the same thing.
  6. Then animate, and test it the third time, not the first.
Five questions before any animation, in order. Most motion that annoys people fails the first two. Based on NN/g's questions on goal and frequency, Apple's Human Interface Guidelines and WCAG 2.3.3.Diagram: Marcus Andrey, CC BY 4.0.

Ask them in order and stop at the first that tells you to cut. They come from three sources that agree with each other. NN/g asks for the goal and the frequency before any animation is designed (Aurora Harley, Animation for Attention and Comprehension, NN/g, 2014). Apple's guidelines say: "Don't add motion for the sake of adding motion", avoid motion on interactions that happen often, and let people cancel it (Apple, Human Interface Guidelines: Motion). WCAG asks that motion triggered by an interaction can be turned off (W3C, Understanding 2.3.3).


2. Judge it the third time

Now do it five times in a row.
Menu800 ms, with a bounceFun once. By the third time, you're waiting for it
Menu150 ms, fade and a short slideYou barely see it, and the menu is simply there
The same menu, two animations. The slow one is the kind that's nice in a review and tiring in real use, because a menu opens dozens of times a day. Live CSS: 800ms cubic-bezier(0.34, 1.56, 0.64, 1) against 150ms ease-out.

Open both menus once: the bouncy one is more fun. Now open them five times in a row. By the third, you're waiting for the slow one.

That's the most common way motion fails, and you'll never see it in a design review, where everyone watches each animation once. NN/g hears it in usability tests over and over: "this [animation] was nice the first time, but now it's getting annoying" (Harley, NN/g, 2014). Their rule of thumb: the more often people meet an animation, the shorter and more subtle it should be (Page Laubheimer, Executing UX Animations, NN/g, 2020).

A useful split:

How often Examples How much motion
Once First launch, an achievement, finishing a big task Room for character
A few times a session Opening a page, a dialog, a sheet Short and clear, 200 to 300 ms
All the time Menus, toggles, tabs, typing, list items Barely there, or none

3. Don't make people wait

An animation that plays after the click and has to finish before you can do anything is a small wait you designed on purpose. Willenskomer calls these non-realtime interactions: they have "the effect of briefly locking the user out" until the transition completes (Issara Willenskomer, The UX in Motion Manifesto, 2017). Apple's rule is the fix: "don't make people wait for an animation to complete before they can do anything" (Apple HIG: Motion).

In practice: people should be able to tap the next thing while a transition is still running, and the interface should follow them, not finish its show first. Untick any demo in this lesson halfway: CSS transitions turn around from wherever they are. That's the minimum.

Motion also doesn't make slow systems fast. NN/g notes that animations used as "time-filling" during transitions are the ones that often frustrate people in usability tests (Laubheimer, The Role of Animation and Motion in UX, NN/g, 2020). If the wait is real, show progress honestly; that's Feedback while waiting.


4. Motion that makes people unwell

For some people, motion isn't annoying: it's physically harmful. W3C explains that movement on screen can trigger vestibular (inner ear) disorder reactions, from dizziness and nausea to migraines, and "potentially needing bed rest to recover" (W3C, Understanding 2.3.3). The usual triggers are large movements: parallax, zooming, things sliding across the whole screen, scrolling that moves more than the content.

Two WCAG rules cover motion directly, and Step 6 goes deeper:

  • 2.2.2 Pause, Stop, Hide (Level A). Anything that moves on its own, lasts more than five seconds and sits next to other content, like a carousel or a ticker, needs a way to pause, stop or hide it (W3C, Understanding 2.2.2).
  • 2.3.3 Animation from Interactions (Level AAA). Motion triggered by an interaction can be turned off, unless it's essential (W3C, Understanding 2.3.3).

The easiest way to meet 2.3.3 is to respect the reduce motion setting in the operating system, which the browser exposes as prefers-reduced-motion (MDN, prefers-reduced-motion). Every demo in this course does. This is the rule that turns the slides in this lesson into fades:

@media (prefers-reduced-motion: reduce) {
  .mo-a-demo .mo-a-dep, .mo-a-demo .mo-a-menu, .mo-a-demo .mo-a-pop, .mo-a-demo .mo-a-tip, .mo-a-demo .mo-a-item { transform: none; transition: opacity 150ms ease-out; transition-delay: 0ms; }
}

Reduced doesn't mean nothing. The menu still appears, with a short fade, so the change is still visible. More in Vestibular disorders and reduced motion and Designing a reduced-motion version.

And one more from Apple: never make motion the only way to say something important (Apple HIG: Motion). If the animation is off, the state should still be clear.


5. Places motion rarely belongs

  • The edge of the screen while people read. A chat bubble sliding in, a "back to top" button flying up. Movement in the periphery pulls the eye away from the task (Harley, NN/g, 2014). If it must appear, fade it in; better, keep it there all the time.
  • Scroll effects. Parallax and scroll-jacking are "frequently frustrating, dizzying, and annoying" (Laubheimer, NN/g, 2020), and they're vestibular triggers. See Scroll and motion.
  • Things that loop forever. A pulsing button, a moving background. They never stop asking for attention.
  • Dense work screens. In a table people scan for hours, a row that animates on every update is noise. A brief highlight that fades is enough.

6. A cut is a design

An instant change is not a failure to animate. Nielsen's response-time limits say that at about 0.1 second, people feel the system reacts instantly and "no special feedback is necessary except to display the result" (Jakob Nielsen, Response Times: The 3 Important Limits, NN/g, 1993). Switching tabs, sorting a column, typing in a field: show the result, now.

The test is simple: if you removed this animation, would anyone lose track of what happened? If not, cut it, and spend the motion where it does a job.


7. Common mistakes

  • Reviewing motion by watching each animation once, in a prototype, on a big screen.
  • Long, showy animations on things people use dozens of times a day.
  • An animation that blocks the next tap until it finishes.
  • Carousels and tickers that move on their own with no pause.
  • Ignoring the reduce motion setting, or treating it as "turn everything off" and losing the information.
  • Using a spinner animation to make a slow system feel faster.

8. Vocabulary

Term Meaning
Frequency How often one person meets an animation in a session
Non-realtime motion Motion that plays after an action and can block the next one
Periphery The edges of the view, where motion grabs attention hardest
Vestibular disorder An inner-ear condition; motion on screen can cause dizziness and nausea
prefers-reduced-motion The CSS media feature that tells a page the person asked for less motion
Cut An instant change from one state to another, with no animation

9. Try it (no tools needed)

  1. Pick the animation in your product that people meet most often.
  2. Trigger it ten times in a row, as fast as you can.
  3. Then turn on reduce motion (on a Mac: System Settings, Accessibility, Display; on an iPhone: Settings, Accessibility, Motion) and do it again.

Reflection questions: At which repetition did it start to feel slow? With reduce motion on, did the change still make sense, or did the information disappear with the motion?


10. If you already practise

  • Add "how often" and "reduced version" as fields in every motion spec, next to duration and easing.
  • Review motion on a real device, repeated, with reduce motion both on and off.
  • Make "no animation" a valid option in your design system, documented with the reasons. It saves the same debate in every review.

Checklist

Learn more

Books, courses and videos to go deeper

Book links go to Amazon. As an Amazon Associate I earn from qualifying purchases, at no extra cost to you.

What motion does in an interface · Vestibular disorders and reduced motion · WCAG and motion · Feedback while waiting · Accessibility fundamentals · Motion purpose check

Saved only on this device.

↑ ↓ to move · Enter to open · Esc to close